Started playing with GPT3, and building services on top of it. I built what I believe is a compelling service (unfortunately if I monetize it, it'll probably be replaced by openAI easily)
If it is, you could get acquihired, or build an even better thing on a competing service in the future, if enough people build things on openai only to be replaced internally they'll stop supporting the api/using it and other competitors will fill that void, but as much as Microsoft is investing I think they will keep it going for indie devs because of github/tooling services they offer that they make a lot of $$ off.